ez-transforms

Tech Stack

• React (TypeScript)
• Tailwind, NextUI, Monaco Text Editor, Axios
• Flask (Backend API)

About

Improving the official albumentations-demo

• Upload your image and select your transform(s) from the albumentations library
• Discover the relationships between transforms and finetune your data augmentation pipeline
• Edit transform properties to experiment with hyperparameter tuning
• Generate different variations of the transforms side-by-side for comparison and download your results

Usage

The backend service is hosted on AWS. I don't plan on maintaining a domain name so the frontend hosted on Vercel won't be able to serve HTTP requests to generate the transformed image. To run the application locally:

1. Clone this repository git clone https://github.com/teddyld/ez-transforms.git

2. Run the frontend locally

   cd frontend
   npm install
   npm run dev
   

3. Optionally run the backend locally

   cd backend
   pip install requirements.txt
   python3 app.py
   

Support

The version of albumentations used is 1.4.6.